Design of Clinical Trials for Mild to Moderate Ulcerative Colitis

Most patients with ulcerative colitis (UC) have mild to moderate disease, for which 5-aminosalicylates (5-ASAs) have remained the mainstay of treatment because of their efficacy, low cost, and excellent safety profile.1Feuerstein J.D. Isaacs K.L. Schneider Y. et al.AGA clinical practice guidelines on the management of moderate to severe ulcerative colitis.Gastroenterology. 2020; 158: 1450-1461Google Scholar, 2Harbord M. Eliakim R. Bettenworth D. et al.Third European evidence-based consensus on diagnosis and management of ulcerative colitis. Part 2: current management.J Crohns Colitis. 2017; 11: 769-784Google Scholar, 3Ordás I. Eckmann L. Talamini M. et al.Ulcerative colitis.Lancet. 2012; 380: 1606-1619Google Scholar However, a substantial proportion of patients fail to maintain remission with 5-ASAs alone and require additional therapy.3Ordás I. Eckmann L. Talamini M. et al.Ulcerative colitis.Lancet. 2012; 380: 1606-1619Google Scholar To date, 5 classes of advanced treatments have been approved for the management of moderately to severely active UC: tumor necrosis factor antagonism, integrin antagonism, interleukin-12/23 blockade, Janus kinase inhibition, and sphingosine-1-phosphate receptor modulation. Although these agents are safe and effective, there remains a therapeutic gap for patients with mildly to moderately active UC for whom 5-ASAs fail or who are intolerant to 5-ASAs yet do not meet the criteria for a biologic or novel small-molecule therapy. Hence, there has been a resurgence of interest in developing treatments for mildly to moderately active UC, particularly with microbiome-targeted, gut-restricted, and/or nonconventional complementary medicines. Almost all recent randomized controlled trials (RCTs) have recruited patients with moderately to severely active UC,3Ordás I. Eckmann L. Talamini M. et al.Ulcerative colitis.Lancet. 2012; 380: 1606-1619Google Scholar for which regulatory guidance has recommended appropriate efficacy assessments, endpoint definitions, and timing to measure these outcomes.4European Medicines AgencyGuideline on the development of new medicinal products for the treatment of ulcerative colitis.https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-development-new-medicinal-products-treatment-ulcerative-colitis-revision-1_en.pdfGoogle Scholar,5US Food and Drug AdministrationUlcerative colitis: clinical trial endpoints: guidance for industry.https://www.fda.gov/files/drugs/published/Ulcerative-Colitis--Clinical-Trial-Endpoints-Guidance-for-Industry.pdfGoogle Scholar An important distinction is that clinical trials enroll patients on the basis of disease activity rather than severity. There are unique considerations for trial designs enrolling patients with milder activity. First, there are no consensus definitions for mild to moderate disease activity, resulting in heterogeneous study populations.6Singh S. Chowdhry M. Umar S. et al.Variations in the medical treatment of inflammatory bowel disease among gastroenterologists.Gastroenterol Rep (Oxf). 2018; 6: 61-64Google Scholar Second, placebo rates are higher in mild disease activity, and current endpoints used to classify therapeutic efficacy may be insensitive to change.7Sedano R, Hogan M, Nguyen TM, et al. Systematic review and meta-analysis: clinical, endoscopic, histologic and safety placebo rates in induction and maintenance trials of ulcerative colitis. J Crohns Colitis. Published July 26, 2021. https://doi.org/10.1093/ecco-jcc/jjab135.Google Scholar Third, patient-reported outcomes and objective measures are poorly correlated in mildly active UC, yet precise endoscopic and histologic targets remain unclear.8Ma C. Sandborn W.J. D’Haens G.R. et al.Discordance between patient-reported outcomes and mucosal inflammation in patients with mild to moderate ulcerative colitis.Clin Gastroenterol Hepatol. 2020; 18: 1760-1768Google Scholar,9Ma C. Sedano R. Almradi A. et al.An international consensus to standardize integration of histopathology in ulcerative colitis clinical trials.Gastroenterology. 2021; 160: 2291-2302Google Scholar Given these challenges, clarity is needed regarding the optimal trial design for mildly to moderately active UC. We summarize the advantages and limitations of current study designs, outline areas of research priority, and propose strategies to improve the efficiency and validity of RCTs for this population. We aim to provide expert, opinion-based recommendations for trial designs for mildly to moderately active UC. However, the proposed strategies in this article have been informed by a systematic review conducted in ClinicalTrials.gov up to April 8, 2021, using terms to identify mild or mildly to moderately active UC trials enrolling adult patients. We excluded studies of behavioral or surgical interventions and/or those aiming to modify parameters other than disease activity (eg, fatigue). We screened 224 trials; 116 did not meet the inclusion criteria, and 47 did not define a population with mildly to moderately active UC. A total of 61 studies were included (Supplementary Table 1): 38 trials (62%) were placebo controlled, 11 (18%) were active comparator studies, and 11 (18%) were uncontrolled. A broad range of interventions were assessed, including 5-ASAs, biologics, small molecules, corticosteroids, diet, herbals, and fecal transplant. There are no standardized definitions of mildly to moderately active UC. Multiple society guidelines endorse the Truelove and Witts criteria,1Feuerstein J.D. Isaacs K.L. Schneider Y. et al.AGA clinical practice guidelines on the management of moderate to severe ulcerative colitis.Gastroenterology. 2020; 158: 1450-1461Google Scholar,10Mowat C. Cole A. Windsor A. et al.Guidelines for the management of inflammatory bowel disease in adults.Gut. 2011; 60: 571-607Google Scholar although this tool predominantly identifies systemically unwell patients with severe disease rather than patients with mild activity. Most trials use the Mayo Clinic Score (MCS) or near-equivalent Ulcerative Colitis Disease Activity Index (UCDAI). These composite measures include a rectal bleeding subscore (RBS), stool frequency subscore (SFS), endoscopic appearance, and physician global assessment (PGA). In our review, 50 studies used the MCS for inclusion (Table 1), yet 20 different cutoffs were used to define mildly to moderately active UC, with minimum scores of 1–6 and maximum scores of 4–11. In contrast, moderately to severely active UC is typically defined by an MCS of 6–12.11Ma C. Panaccione R. Fedorak R.N. et al.Heterogeneity in definitions of endpoints for clinical trials of ulcerative colitis: a systematic review for development of a core outcome set.Clin Gastroenterol Hepatol. 2018; 16: 637-647Google Scholar It should be recognized that the MCS was developed in the 1980s, when the science of index development and item response theory was still in its infancy. Limitations of MCS-based definitions for mild to moderate disease activity include the following:•Patients may meet the inclusion threshold despite having no endoscopic inflammation.•Patients with moderate disease at the higher end of the range are much sicker than those with mild disease at the low end of the spectrum.•Patients may be qualified using the PGA, which is no longer recommended by regulatory authorities.•Potentially important symptoms to patients, such as fecal urgency, are not included.Table 1Disease Activity and Endpoint Definitions in Contemporary Clinical Trials in Mild to Moderate Ulcerative ColitisTrial identification numberIntervention/comparatorMild/moderate definitionClinical responseClinical remissionEndoscopic responseEndoscopic remissionMucosal healingHistologic remissionNCT00628433HE3286/placeboMCS = 4-9NANANANANANANCT01320436Curcumin/5-ASASCCAI of >5 and ≤13SCCAI of ≤2 or drop of ≥3 pointsSCCAI of ≤2NANANANANCT03861143BT-11/placeboMCS = 4–10 at baseline with an MES = 2 (confirmed by central reader)NAMCS ≤2 (RBS = 0, SFS = 0, MES = 0–1)NAMES = 0–1MES = 0–1 and Geboes score < 3.1NANCT00063830Alicaforsen/placeboUCDAI = 4–10 including abnormal endoscopic scoreNANANANANANANCT02093767Prebiotic preparation containing inulin and oligofructose/NAMCS = 3-8MCS ≤2 (RBS = 0, SFS = 0, MES = 0–1)NAMES = 0–1NANANCT04034758FMT by standardized quantitative multidonor intestinal microbiota capsule/placeboMCS = 4–10 and MES ≥ 2MCS decrease of ≥3 points and ≥30% from baseline and decrease of ≥1 point in RBS or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1 point)NANANANANCT01882764HMPL-004/placeboMCS = 4–10 and MES = 2–3NAMCS ≤ 2 (no subscore > 1 point and RBS = 0)NANANANANCT01164904Amg181/NAMCS = 4–9 and MES > 1NANANANANANANCT02488954Probiotics via portion of cheese (Propionibacterium freudenreichii)/NASCCAI = 4–11SCCAI decrease of ≥2 points from baselineSCCAI < 3NANAMES = 0–1NANCT01045018Mesalamine delayed release tablets/placeboUCDAI = 4–9 with PGA = 1–2 and mucosal appearance score = 1–2NANANANANANANCT02683733Bioenhanced curcumin/5-ASAMCS = 2–6NAMCS = 0–1NAUCEIS < 3NANANCT01149707PUR 0110 rectal enema/placeboMCS = 5–10, MES ≥ 2, and RBS ≥ 2MCS decrease of ≥3 points from baselineMCS ≤ 4 (MES ≤ 1, RBS = 0, SFS ≤ 3)NANANANANCT00457171GI-270384/placeboMTWI = 4–11 with an endoscopy score of 2NANANANANANANCT03006809FMT with or without vancomycin/NAMCS = 4–9, MES ≥ 1MCS decrease of ≥3 points from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2MES decrease of ≥1 point from baselineNAMES = 0–1NANCT00548574MMX mesalazine (SPD476)/AsacolUCDAI 4–10 (inclusive) with a sigmoidoscopy score ≥ 1 and PGA ≤ 2NAUCDAI ≤ 1 (RBS = 0, SFS = 0)NANANANANCT03773952PBF-677/placeboMCS ≤ 6, RBS ≤ 2, and/or BFS ≤ 2.NANANANANANANCT00659802HMPL-004/placeboMCS 4–10, with activity confirmed by endoscopy within 2 weeks before study entry, with MES ≥ 1MCS decrease of ≥3 points and >30% from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1)NANAMES ≤ 1NANCT04102852LGG ATCC 53103/LGG regular doseMCS = 2–4NANANANANANANCT00951548VSL#3/placeboUCDAI = 3–8NANANANANANANCT04074590LYS006/placeboMCS = 5–10, with MES = 2–3, RBS = 1–3, BFS = 1–4, and PGA = 1–2NAMCS ≤ 2 (no subscore > 1)NANANANANCT00867438Parnaparin sodium (CB-01-05-MMX) and 5-ASA/placebo and 5-ASAUCDAI of ≥4 and ≤10, CAI of ≥5 and ≤12NASCCAI < 4NANANANANCT01039597ORE1001/placeboBaron score ≥ 2, Truelove-Witt (modified) ≤ 14NANANANANANANCT00603733Oral extended-release Pentasa (mesalamine)/PentasaUCDAI = 3–8 and a score of ≥1 for endoscopyUCDAI decrease of ≥2 points from baselineUCDAI ≤ 2NANANANANCT01903252TP05/AsacolMCS ≥ 5, sigmoidoscopy component score of ≥2 confirmed by central review, and RBS ≥ 1MCS decrease of ≥3 points and >30% from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2MES decrease of ≥1 point from baselineMES = 0NANANCT01567956Propionyl-l-carnitine/placeboUCDAI = 3–6, with RBS ≥ 1NAMCS ≤ 2NAMES ≤ 1NANANCT02246686STW5-II (Iberogast N, BAY98-7410)/placeboCAI ≥ 5–10NASCCAI ≤ 4NANANANANCT01562314GWP42003/placeboMCS of ≥4 and ≤10 with MES ≥ 1NANANANANANANCT03716388FMT/placebo or mesalamineMCS = 4–10, MES > 1MCS decrease of ≥3 points and >30% from baselineMCS ≤ 2 (all subscores ≤ 1)NAMES = 0NANHI = 0–1NCT02808390GED-0507-34-Levo/placeboMCS of ≥4 to ≤8NANANANANANANCT00408174Balsalazide disodium/mesalazineMCS = 6–10 with RBS ≥ 2 and MES ≥ 2NANANANANANANCT01453491SRT2104/NAMCS = 6–10 with RBS ≥ 1, MES = 2–3, and PGA < 3NANANANANANANCT02201758Flaxseed lignan-enriched complex/placeboMCS = 3–10MCS decrease of ≥3 points from baselineNANANANANANCT00679380Budesonide-MMX/placebo or Entocort ECUCDAI of ≥ 4 and ≤ 10UCDAI decrease of ≥3 points from baselineUCDAI ≤ 1 (RBS = 0, SFS = 0)MES decrease of ≥1 point from baselineNANANANCT01036022GSK1399686/placebo or mesalazineUCDAI = 4–10 with RBS ≥ 1, endoscopy score ≥ 1, and PGA < 3SCCAI decrease of ≥2 points from baselineSCCAI < 3NANANANANCT04018040Lacto-ovo vegetarian diet/NAMCS ≤ 9MCS decrease of ≥3 points from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1)NANANANANCT03123120Spesolimab/placeboMCS ≤ 10NAMCS ≤2 (no subscore > 1)NANAMES ≤ 1RHI ≤ 6NCT02390726FMT/placeboMCS = 4–10, MES = 1–2NANANANANANANCT01026857Propionyl-l-carnitine/ placeboUCDAI = 3–10NANANANANANANCT02368717Mesalazine enema/placeboMCS = 4–10, MES ≥ 2MCS decrease of ≥3 points and >30% from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1)NANANANANCT01612039ASP3291/placeboModified Baron Score ≥ 2 and Ulcerative Colitis Clinical Score ≥ 4NANANANANANANCT03843385Encapsulated FMT/placeboMCS = 4–9 and MES > 1MCS decrease of ≥3 points from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1)NANANANANCT03998488FMT and psyllium husk powder/placebo with or without psyllium huskMCS = 4–10 with MES ≥ 1MCS decrease of ≥3 points and >30% from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1)MES = 0–1 or decrease of ≥1 point from baselineMES = 0NANANCT00751933Vivotif and Dukoral vaccination with or without oats/placeboSCCAI of >5 and ≤13NANANANANANANCT04401605Fermented food–supplemented diet/placeboMCS = 2–7MCS decrease of ≥3 points from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 2 (no subscore > 1)NANANANANCT02493712IBD98-M/placeboUCDAI ≥ 4 and ≤10 and endoscopic subscore ≥ 1UCDAI decrease of ≥3 points from baselineUCDAI ≤ 1 (RBS = 0, SFS = 0)NANANANANCT00793130Coltect (curcumin, green tea, and selenomethionine)/NACAI of >4 and ≤8NANANANANANANCT04398550Specific carbohydrate diet/Mediterranean dietMCS = 3–11, MES ≤ 2, or fecal calprotectin > 150 μg/gNANANANANANANCT01506362BL-7040/NAMCS of ≥5 and ≤9, MES ≥ 2, and RBS ≥ 1MCS decrease of ≥3 points from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1NANANAMES decrease of ≥1 point from baseline and MES ≤ 1NANCT03521232Niclosamide/NAMCS of ≥4 to <8NAMCS ≤ 2 (no subscore > 1)NANANANANCT02361957Ecologic 825 (probiotic)/placeboSCCAI < 5NANANANANANANCT00065065Rosiglitazone/placeboUCDAI = 4–10NAMCS ≤ 2NANANANANCT01059344Mesalazine/placeboUCDAI = 4–10 with sigmoidoscopy component score ≥ 2 and RBS ≥ 1UCDAI decrease of ≥3 points from baselineUCDAI ≤ 1NAUCDAI endoscopic subscore ≤ 1NANANCT02058524FMT/NASCCAI > 4–9NANANANANANANCT03110289Superdonor FMT/autologous FMTMCS = 4–10, MES = 2–3MCS decrease of ≥3 points from baselineMCS ≤ 2 (all subscores ≤1)MES decrease of ≥1 point from baselineNANANANCT03986996Amoxicillin and doxycycline/amoxicillin, metronidazole, and doxycyclineSCCAI of ≥5 and ≤10SCCAI decrease of >3 points from baselineSCCAI < 5NANANANANCT00063414Alicaforsen/mesalamine enemaUCDAI = 4–10NANANANANANANCT01538251Propionyl-l-carnitine/ placeboUCDAI = 3–6 with RBS ≥ 1NAMCS ≤ 2 (RBS = 0, no subscore > 1)NANANANANCT01100112Budesonide-MMX/NAUCDAI ≤ 10UCDAI decrease of ≥3 points from baselineUCDAI ≤ 1 (RBS = 0, SFS = 0)UCDAI endoscopic subscore decrease of ≥1 point from baselineNANANANCT00679432Budesonide-MMX/ placebo or mesalazineUCDAI of ≥4 and ≤10NAUCDAI ≤ 1 (RBS = 0, SFS = 0)UCDAI endoscopic subscore decrease of ≥1 point from baselineNANANANCT01177228Vedolizumab/placeboMCS = 1–7NANANANANANANCT02762500LYC-30937-EC/placeboMCS = 4–11, MES ≥ 2, and RBS ≥ 1MCS decrease of ≥3 points and >30% from baseline with a decrease in RBS of ≥1 point or absolute RBS ≤ 1MCS ≤ 1 (RBS = 0, SFS = 0)NANANANABFS, Bowel Frequency Score; CAI, Clinical Activity Index; FMT, fecal microbiota transplantation; LGG, Lactobacillus rhamnosus GG; MTWI, Modified Truelove/Witts activity index; NA, not applicable; NHI, Nancy Histology Index; SCCAI, Simple Clinical Colitis Activity Score; UCDAI, Ulcerative Colitis Disease Activity Index; UCEIS, Ulcerative Colitis Endoscopic Index of Severity. Open table in a new tab BFS, Bowel Frequency Score; CAI, Clinical Activity Index; FMT, fecal microbiota transplantation; LGG, Lactobacillus rhamnosus GG; MTWI, Modified Truelove/Witts activity index; NA, not applicable; NHI, Nancy Histology Index; SCCAI, Simple Clinical Colitis Activity Score; UCDAI, Ulcerative Colitis Disease Activity Index; UCEIS, Ulcerative Colitis Endoscopic Index of Severity. In trials for moderately to severely active UC, enrolling patients using centrally read endoscopy provides an objective assessment of mucosal inflammation and reduces placebo rates by minimizing observation bias.12Ma C. Guizzetti L. Panaccione R. et al.Systematic review with meta-analysis: endoscopic and histologic placebo rates in induction and maintenance trials of ulcerative colitis.Aliment Pharmacol Ther. 2018; 47: 1578-1596Google Scholar This enrollment method has become a regulatory standard; however, defining mild to moderate endoscopic inflammation is challenging. A Mayo Endoscopic Score (MES) of 1 is considered abnormal but, until recently, was also considered endoscopic remission. (MES = 0/1 is now considered endoscopic improvement.) Furthermore, patients with an MES = 1 are likely a heterogeneous group who may or may not have histologic- or biomarker-based disease activity. Heterogeneous definitions of clinical response and remission are used in contemporary trials for mildly to moderately active UC.11Ma C. Panaccione R. Fedorak R.N. et al.Heterogeneity in definitions of endpoints for clinical trials of ulcerative colitis: a systematic review for development of a core outcome set.Clin Gastroenterol Hepatol. 2018; 16: 637-647Google Scholar Depending on the inclusion criteria used, definitions are based on reductions in MCS (eg, ≥3 points and ≥30% from baseline) or absolute MCS cutoffs (eg, MCS ≤ 2 with no subscore > 1) (Table 1). Although symptoms are nonspecific, the use of objective outcomes was rare among the trials in our review, with endoscopic response or remission measured in 13 of 61 (21%) trials and histologic remission in 2 of 61 (3%) trials. Among the 8 trials that measured endoscopic remission, a stringent definition of MES = 0 was used in 3 of 8 studies, whereas 4 of 8 studies used MES = 0/1. However, the patients in remission with MES = 0/1 may overlap with patients who had mild endoscopic disease at baseline. For histologic endpoints, validated cutoffs confirmed empirically in RCTs are required. The contemporary standard in UC RCTs is to enroll symptomatic patients with confirmed endoscopic inflammation. Clinical assessment alone is inadequate in milder disease activity because symptoms such as stool frequency are nonspecific, and rectal bleeding may be less dominant. Many trials are therefore using an MES cutoff to define the minimum endoscopic activity required for inclusion. Some trials include patients with an MES of ≥2 at baseline, recognizing that this criterion overlaps with moderate to severe disease activity and may lead to ineligibility of patients with milder endoscopic inflammation. However, lowering the endoscopic bar to an MES of ≥1 at enrollment is likely to increase the placebo response rate, reduce trial efficiency, and limit the detection of efficacy signals.7Sedano R, Hogan M, Nguyen TM, et al. Systematic review and meta-analysis: clinical, endoscopic, histologic and safety placebo rates in induction and maintenance trials of ulcerative colitis. J Crohns Colitis. Published July 26, 2021. https://doi.org/10.1093/ecco-jcc/jjab135.Google Scholar One approach to overcoming these challenges is to combine endoscopic assessment with histopathology for trial inclusion, permitting the inclusion of patients with mildly to moderately active UC who have MES = 1 plus evidence of histologic inflammation. A recent consensus suggested that the presence of neutrophilic inflammation, defined by a Geboes score of ≥3.1, Robarts Histopathology Index (RHI) of ≥4, or Nancy index of ≥2, were appropriate minimum disease activity requirements for enrollment.9Ma C. Sedano R. Almradi A. et al.An international consensus to standardize integration of histopathology in ulcerative colitis clinical trials.Gastroenterology. 2021; 160: 2291-2302Google Scholar To the best of our knowledge, no trials to date in mildly to moderately active UC have implemented histology as an inclusion criterion, in part because of the operational complexities of acquiring, processing, and reading pathology slides within a sufficiently rapid turnaround time to allow randomization. However, there are several conceptual advantages to including histology at enrollment. First, if histology is used as an endpoint in clinical trials, then it would be logical to enroll patients with baseline histologic activity. One third of patients in clinical remission with MES = 0/1 have ongoing histologic inflammation, and patients who achieve endoscopic and histologic remission have a 61% lower annual risk of clinical relapse compared to patients with persistent histologic activity.13Park S. Abdi T. Gentry M. et al.Histological disease activity as a predictor of clinical relapse among patients with ulcerative colitis: systematic review and meta-analysis.Am J Gastroenterol. 2016; 111: 1692-1701Google Scholar,14Yoon H. Jangi S. Dulai P.S. et al.Incremental benefit of achieving endoscopic and histologic remission in patients with ulcerative colitis: a systematic review and meta-analysis.Gastroenterology. 2020; 159: 1262-1275Google Scholar Accordingly, the US Food and Drug Administration is encouraging the use of histologic assessment to evaluate mucosal healing as an endpoint in UC clinical trials.5US Food and Drug AdministrationUlcerative colitis: clinical trial endpoints: guidance for industry.https://www.fda.gov/files/drugs/published/Ulcerative-Colitis--Clinical-Trial-Endpoints-Guidance-for-Industry.pdfGoogle Scholar Second, it is plausible that using histology as an entry criterion may improve responsiveness by enriching for patients with active inflammation at enrollment, in a manner analogous to confirming endoscopic activity by central reading. In a 10-week trial of 281 patients with mildly to moderately active UC treated with delayed-release mesalamine 4.8 g/day, Feagan et al15Feagan B.G. Sandborn W.J. D’Haens G. et al.The role of centralized reading of endoscopy in a randomized controlled trial of mesalamine for ulcerative colitis.Gastroenterology. 2013; 145: 149-157Google Scholar could not demonstrate a significant treatment difference for the primary outcome of clinical remission vs placebo when all patients were included (30.0% vs 20.6%; P = .069), yet when patients without active endoscopic disease activity were excluded by central readers, the treatment difference was highly significant (29.0% vs. 13.8%; P = 0.011). A similar study evaluating enrollment by histologic inflammation at baseline should be conducted. An alternative strategy for trial inclusion is to integrate biomarkers as screening tools, which has been previously established in Crohn’s disease trials of natalizumab, certolizumab, and vedolizumab that recruited patients with elevated baseline C-reactive protein and symptoms.16Targan S.R. Feagan B.G. Fedorak R.N. et al.Natalizumab for the treatment of active Crohn’s disease: results of the ENCORE trial.Gastroenterology. 2007; 132: 1672-1683Google Scholar, 17Schreiber S. Rutgeerts P. Fedorak R.N. et al.A randomized, placebo-controlled trial of certolizumab pegol (CDP870) for treatment of Crohn’s disease.Gastroenterology. 2005; 129: 807-818Google Scholar, 18Sandborn W.J. Feagan B.G. Rutgeerts P. et al.Vedolizumab as induction and maintenance therapy for Crohn’s disease.N Engl J Med. 2013; 369: 711-721Google Scholar In UC, fecal calprotectin (FC) is more sensitive than C-reactive protein, is well correlated with endoscopic inflammation, and can accurately differentiate patients with inflammatory disease from those with functional disorders.19Mosli M.H. Zou G. Garg S.K. et al.C-reactive protein, fecal calprotectin, and stool lactoferrin for detection of endoscopic activity in symptomatic inflammatory bowel disease patients: a systematic review and meta-analysis.Am J Gastroenterol. 2015; 110: 802-819Google Scholar The American College of Gastroenterology suggests using an FC cutoff of >150–200 μg/g to identify patients with mildly to moderately active UC, although interindividual variability is significant, and the performance characteristics of the test depend on the threshold used. Given the first-line positioning of 5-ASAs for mildly to moderately active UC, practical entry criteria for future trials include failure or intolerance of 5-ASAs at optimal dosing (≥4 g/day) and use of combination oral and rectal 5-ASAs in patients with left-sided disease. A greater proportion of UC patients are now being exposed to biologic agents, and whether this exposure should be exclusionary for trials of mildly to moderately active UC is controversial. Among the trials included in this review, 4 of 61 (7%) allowed concomitant biologic therapies, and 17 of 61 (28%) had specific requirements pertaining to their washout. These trials may provide important data on the optimal treatment strategy in patients who have had an inadequate or partial response to biologic therapy who may not yet warrant a change in drug class or in patients for whom a biologic agent fails but who do not have severe activity after biologic withdrawal. However, from a practical perspective, regulators are often focused on whether patients have experienced failure of 5-ASAs and are reluctant to include patients for whom biologics failed. Concomitant corticosteroids are typically allowed in trials for moderately to severely active UC , with stipulations on the maximum dose, fixed dosing during induction therapy, forced tapering during maintenance therapy, and use as rescue therapy or reintroduction.20George J. Singh S. Dulai P.S. et al.Corticosteroid-free remission vs overall remission in clinical trials of moderate-severe ulcerative colitis and Crohn’s disease.Inflamm Bowel Dis. 2020; 26: 515-523Google Scholar In trials of mildly to moderately active UC, milder symptoms may be completely masked by higher-dose corticosteroids. Therefore, it would be reasonable to consider excluding all patients on concomitant systemic corticosteroids at baseline. This would simplify the trial design and minimize the risk of confounding treatment efficacy with corticosteroids, especially for symptoms. However, recognizing that some patients with mildly to moderately active UC may require a short course of low-dose corticosteroids to bridge their symptomatic response and that completely excluding patients on corticosteroids may make recruitment more challenging (particularly during the screening period), an alternative would be to establish a lower maximum allowable dose (eg, 15–20 mg/day) during induction. This would require consideration of corticosteroid tapering and reporting of corticosteroid-free remission as an endpoint. During maintenance, the threshold to permit the reintroduction or dose escalation of corticosteroids may be higher in patients with milder UC to limit the possibility of corticosteroid-dependent patients achieving clinical remission. Regulatory guidelines for drug development in UC recommend a randomized, double-blind, parallel-group design with an active and/or placebo comparator.4European Medicines AgencyGuideline on the development of new medicinal products for the treatment of ulcerative colitis.https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-development-new-medicinal-products-treatment-ulcerative-colitis-revision-1_en.pdfGoogle Scholar The chosen comparator depends on the positioning of the investigational drug. First-line therapies for mildly to moderately active UC should be compared against 5-ASAs, the current treatment standard. Placebo is an acceptable comparator against investigational second-line drugs for patients who did not tolerate 5-ASAs or for whom 5-ASAs failed. Clinical remission rates for oral 5-ASAs in UC are approximately 30%, whereas a recent meta-analysis of 84 pooled UC induction trials showed a clinical remission rate of 11% for placebo.7Sedano R, Hogan M, Nguyen TM, et al. Systematic review and meta-analysis: clinical, endoscopic, histologic and safety placebo rates in induction and maintenance trials of ulcerative colitis. J Crohns Colitis. Published July 26, 2021. https://doi.org/10.1093/ecco-jcc/jjab135.Google Scholar,21Murray A. Nguyen T.M. Parker C.E. et al.Oral 5-aminosalicylic acid for induction of remission in ulcerative colitis.Cochrane Database Syst Rev. 2020; 8CD000543Google Scholar These findings highlight the need for much larger samples to demonstrate the superiority in a 5-ASA–controlled trial than in a placebo-controlled study using a frequentist approach. Alternative strategies such as an adaptive design using the uniformly unbiased minimum-variance estimator method for analysis or Bayesian estimates of placebo response may reduce sample size requirements.22Jung S.H. Kim K.M. On the estimation of the binomial probability in multistage clinical trials.Stat Med. 2004; 23: 881-896Google Scholar,23Koyama T. Chen H. Proper inference from Simon’s two-stage designs.Stat Med. 2008; 27: 3145-3154Google Scholar Despite the statistical efficiency of comparing to placebo, ethical considerations must also be evaluated when choosing a comparator arm. This is especially true for agents using mechanisms of action that have already been shown to be effective in other pivotal trials (eg, biosimilars) and where randomizing to placebo may be limited by unclear equipoise. Here, the use of a historical control arm may be more appropriate. Current trials of mildly to moderately active UC vary in duration depending on the design (Supplementary Table 1). Most trials include an induction period, ranging from 2 to 16 weeks (most commonly 8–12 weeks), and a maintenance period, ranging from 36 weeks to 104 weeks (most commonly 52 weeks). The optimal duration will depend on several factors, including the phase of the study and the mechanism of action of the agent under investigation. Generally, consistent achievement of endoscopic and histologic endpoints in studies of moderately to severely active UC has required at least 8–12 weeks, although earlier assessment of symptomatic response is important in patients with active UC. Clinical remission defined by the MCS (SFS = 0, RBS = 0, and MES ≤ 1, with an alternative definition of a ≥1-point decrease from baseline to achieve an SFS = 0/1) remains the primary endpoint in most UC RCTs.5US Food and Drug AdministrationUlcerative colitis: clinical trial endpoints: guidance for industry.https://www.fda.gov/files/drugs/published/Ulcerative-Colitis--Clinical-Trial-Endpoints-Guidance-for-Industry.pdfGoogle Scholar However, in milder UC activity, approximately 25% of asymptomatic patients have an MES of >1, and conversely, up to half of patients with MES = 0 may have persistently increased stool frequency.8Ma C. Sandborn W.J. D’Haens G.R. et al.Discordance between patient-reported outcomes and mucosal inflammation in patients with mild to moderate ulcerative colitis.Clin Gastroenterol Hepatol. 2020; 18: 1760-1768Google Scholar,24Colombel J.F. Keir M.E. Scherl A. et al.Discrepancies between patient-reported outcomes, and endoscopic and histological appearance in UC.Gut. 2017; 66: 2063-2068Google Scholar The US Food and Drug Administration regulatory framework for a mucosal healing label now requires both endoscopic and histologic assessments, which is emerging as a novel coprimary endpoint in UC. Ustekinumab was the first compound with a label for mucosal healing, defining histologic improvement by neutrophil infiltration in <5% of crypts, with no crypt destruction, erosions, ulcerations, or granulation tissue (equivalent to a Geboes score of ≤3.1).25Sands B.E. Sandborn W.J. Panaccione R. et al.Ustekinumab as induction and maintenance therapy for ulcerative colitis.N Engl J Med. 2019; 381: 1201-1214Google Scholar A meta-analysis showed that the annual risk of clinical relapse was 52% lower in patients who achieve MES = 0 than in those with MES = 1 and 61% lower in patients achieving histologic remission than in those with persistent histologic activity.14Yoon H. Jangi S. Dulai P.S. et al.Incremental benefit of achieving endoscopic and histologic remission in patients with ulcerative colitis: a systematic review and meta-analysis.Gastroenterology. 2020; 159: 1262-1275Google Scholar Given that several trials have reported higher treatment effects with histologic outcomes than with endoscopic or clinical outcomes,14Yoon H. Jangi S. Dulai P.S. et al.Incremental benefit of achieving endoscopic and histologic remission in patients with ulcerative colitis: a systematic review and meta-analysis.Gastroenterology. 2020; 159: 1262-1275Google Scholar,26Feagan B.G. Danese S. Loftus Jr., E.V. et al.Filgotinib as induction and maintenance therapy for ulcerative colitis (SELECTION): a phase 2b/3 double-blind, randomised, placebo-controlled trial.Lancet. 2021; 397: 2372-2384Google Scholar the use of a composite endpoint including histopathology may be more sensitive to subtle changes after treatment in mildly to moderately active UC. FC is widely used in clinical practice, although its ability to support labeling claims remains unclear, and its use as a secondary trial endpoint will likely continue in the immediate future.19Mosli M.H. Zou G. Garg S.K. et al.C-reactive protein, fecal calprotectin, and stool lactoferrin for detection of endoscopic activity in symptomatic inflammatory bowel disease patients: a systematic review and meta-analysis.Am J Gastroenterol. 2015; 110: 802-819Google Scholar In a post hoc analysis of the MOMENTUM trial (Ability to Maintain or Achieve Clinical and Endoscopic Remission With MMX Mesalamine Once Daily in Adults With Ulcerative Colitis) in patients with mildly to moderately active UC treated with mesalamine, FC accurately discriminated patients achieving endoscopic and histologic remission, with the optimal FC cutoffs at weeks 8 and 52 for endoscopic healing being 251 μg/g and 99 μg/g, respectively.27Stevens T.W. Gecse K. Turner J.R. et al.Diagnostic accuracy of fecal calprotectin concentration in evaluating therapeutic outcomes of patients with ulcerative colitis.Clin Gastroenterol Hepatol. 2021; 19: 2333-2342Google Scholar For histologic remission, optimal FC cutoffs at weeks 8 and 52 were found to be 75 μg/g and 99 μg/g, respectively. The advantages of incorporating FC into clinical trials include its wide dynamic range, noninvasiveness, and translatability to clinical practice. Advances in computer processing and artificial intelligence (AI) methods have been increasingly used to automate video and image analysis, with potentially promising applications for future UC trials. Accurate AI algorithms could be used to increase the efficiency of central reading for histopathology and endoscopy endpoints. Several groups have demonstrated that deep learning models can be taught to grade endoscopic disease activity in patients with UC.28Gottlieb K. Requa J. Karnes W. et al.Central reading of ulcerative colitis clinical trial videos using neural networks.Gastroenterology. 2021; 160: 710-719Google Scholar, 29Bossuyt P. Nakase H. Vermeire S. et al.Automatic, computer-aided determination of endoscopic and histological inflammation in patients with mild to moderate ulcerative colitis based on red density.Gut. 2020; 69: 1778-1786Google Scholar, 30Stidham R.W. Liu W. Bishu S. et al.Performance of a deep learning model vs human reviewers in grading endoscopic disease severity of patients with ulcerative colitis.JAMA Netw Open. 2019; 2e193963Google Scholar Gottlieb et al28Gottlieb K. Requa J. Karnes W. et al.Central reading of ulcerative colitis clinical trial videos using neural networks.Gastroenterology. 2021; 160: 710-719Google Scholar applied a machine learning algorithm using convolutional neural networks to predict the MES from videos collected in the phase 2 mirikizumab trial. The authors demonstrated that machine-read endoscopy had excellent agreement with human central readers, including very high specificity for MES = 0 (96.6%) and MES = 1 (92.0%), where subtle differences can cause interobserver reader variability. AI may also be able to detect additional information from endoscopic videos that is challenging to detect with the human eye: for instance, Bossuyt et al31Bossuyt P. De Hertogh G. Eelbode T. et al.Computer-aided diagnosis with monochromatic light endoscopy for scoring histologic remission in ulcerative colitis.Gastroenterology. 2021; 160: 23-25Google Scholar recently developed a real-time automated evaluation of mucosal pericryptal vascular structures using monochromatic light endoscopy, which was translated to an evaluation of histologic activity with high accuracy. Automating machine learning algorithms to further read histology slides would allow for higher throughput, particularly at induction, where rapid turnaround times for randomization are required.32Hagendorn E. Karsen S. Pai R. et al.P137 preliminary validation of a multi-stage machine learning algorithm to assess histological inflammation in inflammatory bowel disease.J Crohns Colitis. 2021; 15: S224-S225Google Scholar The efficient development of novel therapies for patients with mildly to moderately active UC will fill a substantial therapeutic gap. We highlight considerations for disease assessment, patient eligibility, comparator arms, and outcome definitions in this population and summarize our expert recommendations in Figure 1, recognizing that additional empiric data are required to delineate the optimal study design and that different designs may be required depending on the mechanism of action of the agent under consideration. Nevertheless, similar endpoint definitions should be considered, irrespective of therapeutic mechanism, to minimize heterogeneity in outcome reporting and ensure validity of future indirect treatment comparisons. These expert recommendations are primarily based on the MCS, recognizing that this instrument has been used for all recent drug development in UC, although it has substantive limitations and was not constructed using modern index development methods.